3

I've been working recently replicating this paper: "Constructing Street Networks from GPS trajectories". They got a repository with the implemented code. Anyway, the issue is that their data is in the following format:

391821.292854,5821122.049075

Reading carefully the code and the paper I have figured out that these coordinates are in meters after Mercator projection and are probably in ESRI format.

The problem that I have is that I cannot figure it out how to plot them correctly on QGIS. The file that I have is on txt. I'm using the Delimited Text plug in to import the data and playing with the CRS configuration without results. The data appears nowhere near Berlin (where the data is from)

Ian Turton
  • 81,417
  • 6
  • 84
  • 185
Nicolas Falconi
  • 45
  • 3
  • 2
    It is in CSV or Tab delimited text format. – Ian Turton Oct 18 '19 at 16:13
  • 1
    You need to figure out what coordinate reference system the coordinates are in. You're partway there in knowing that it's a mercator projection with the units in meters, but that's not quite enough information. Look in the metadata and see if you can find an EPSG code, it would look something like EPSG:3857. An EPSG code is a way of unambiguously defining the coordinate reference system. – csk Oct 18 '19 at 16:27
  • 3
    It looks like UTM. Try EPSG:32633 which is WGS84 UTM 33 North. It might be on ETRS89 instead (3045), but they're equal within about a meter. – mkennedy Oct 18 '19 at 17:21
  • Thank you so much!!!! That was exactly the one I needed, how did you know??? – Nicolas Falconi Oct 18 '19 at 20:46
  • There isn't really anything called "Esri format" though they have created many formats over the years. – Vince Oct 18 '19 at 23:15

1 Answers1

3

There are two points here:

1. File format The original files are space-separated text files (i.e. similar to csv, but with a space instead of a comma acting as a separator). The example you mention (391821.292854,5821122.049075) is a plain CSV instead.

2. Projection These files seem to be UTM, and since proper metadata are missing, I would infer the correct EPGS code from the file names:

  • Athens -> UTM zone 34N, maybe EPSG:4815 or EPSG:32635
  • Berlin -> UTM zone 33N
  • Chicago -> UTM zone 16N (EPSG:32616)

A quick check with a reliable map (Google Satellite will do) will confirm whether the choice is the right one ;)

Just activate the imported CSV layer and use a Google Satellite map as backdrop: if the points are where you expect them to be, to some degree it is a validation that the EPSG Code/UTM Zone is the correct one.

If, on the other hand, your waypoints are in the middle of an ocean or run over buildings, it is a good hint that you need to revise the projection parameters :D

Please not that this "check" is not a real validation, just a way to exclude some obvious mistakes in the properties of the layer.

RafDouglas C. Tommasi
  • 6,479
  • 1
  • 15
  • 41